diff --git a/README.md b/README.md
index 9e708eb..6af9b1a 100644
--- a/README.md
+++ b/README.md
@@ -202,13 +202,13 @@ The following sub-commands are available.
 
 * Use `awk '{print $2, $1}'` to swap columns. Use `paste` to concatenate series.
 
-## Categorical data
+### Categorical data
 
 * With gne datamash, you can manage to handle categorized data. 
   * `cat test/fixtures/iris.csv | sed '/^$/d' | datamash --header-in --output-delimiter=: -t, -g5 collapse 3,4 | cut -f2-3 -d: | sed 's/:/\n/g' | uplot s -d, -T --fmt xyxy`
   * This is not so easy...
 
-## Time series
+### Time series
 
 * Not yet supported.